Fig. 7 | Scientific Reports

Fig. 7

From: Preoperative MRI-based radiomics analysis of intra- and peritumoral regions for predicting CD3 expression in early cervical cancer

Fig. 7

Assessment of the relative influence of model variables in the SVM algorithm, based on SHAP values. Feature0: wavelet-HH_glszm_GrayLevelNonUniformityNormalized_ADC. Feature1: log-sigma-5–0-mm-3D_firstorder_InterquartileRange_ADC. Feature2: lbp-2D_firstorder_Kurtosis_ADC. Feature3: wavelet-HH_glrlm_LongRunEmphasis_ADC. Feature4: wavelet-LH_glcm_Correlation_ADC. Feature5: wavelet-HL_gldm_DependenceVariance_ADC_3mm. Feature6: gradient_glcm_MCC_ADC. Feature7: lbp-2D_glcm_JointEntropy_ADC. Feature8: lbp-2D_firstorder_InterquartileRange_ADC. Feature9: wavelet-LH_gldm_LargeDependenceLowGrayLevelEmphasis_ADC. Feature10: gradient_glcm_Imc2_ADC. Feature11: log-sigma-5–0-mm-3D_glcm_InverseVariance_T1C. Feature12: exponential_firstorder_Minimum_T2W. Feature13: log-sigma-5–0-mm-3D_glcm_DifferenceVariance_T1C. Feature14: logarithm_glszm_HighGrayLevelZoneEmphasis_T2W. Feature15: lbp-2D_firstorder_Skewness_T1C. Feature16: wavelet-HH_ngtdm_Busyness_ADC. Feature17: lbp-2D_gldm_DependenceVariance_T1C.

Back to article page